Agent-to-Agent Commercial Negotiation Protocol MCP server. Exposes A2CN negotiation capabilities as MCP tools, enabling any MCP-compatible agent framework to negotiate commercial terms autonomously.
A2CN MCP server defines 6 tools with complete input schemas and descriptions. Tool names follow verb_noun convention (a2cn_discover, a2cn_initiate_session, a2cn_send_offer, a2cn_accept, a2cn_reject, a2cn_get_session_status). All parameters have type definitions and descriptions. However, descriptions are minimal (10-50 chars), lacking context on WHEN to use each tool, prerequisites, or recovery guidance. No output schemas are documented, LLMs cannot predict response structure for chaining. Error handling is absent from tool definitions. Risk annotations (READ_ONLY, WRITE, IRREVERSIBLE) are present but not formalized as tool annotations in the MCP schema. Parameter descriptions lack format constraints, ranges, or examples of valid values.
Accept current counterparty offer
Check if counterparty supports A2CN
Poll for counterparty response
Start a negotiation session
Reject and end session
Send a counteroffer
Output schemas not documented. LLMs cannot predict response structure (fields, types, pagination) for downstream tool chaining or data extraction.
Tool descriptions are too brief (10-50 chars). They state WHAT the tool does but omit WHEN to use it, prerequisites, or how it relates to other tools. LLMs cannot reliably select the right tool.
No error handling guidance in tool definitions. Irreversible operations (a2cn_accept, a2cn_reject) lack confirmation or dry-run support. Agents cannot recover from mistakes.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | D | 59 | 2026-07-28+ | v2 |
Parameter descriptions lack format constraints, ranges, or valid value examples. E.g., 'deal_type' accepts 'saas_renewal, goods_procurement, services_engagement, logistics_rate' but this is not formalized as an enum.
Risk annotations (READ_ONLY, WRITE, IRREVERSIBLE) are metadata but not integrated into MCP tool annotations (readOnlyHint, destructiveHint, idempotentHint). Clients cannot parse risk from the schema.